CN102209041B - Scheduling method, device and system - Google Patents
Scheduling method, device and system Download PDFInfo
- Publication number
- CN102209041B CN102209041B CN201110196170.8A CN201110196170A CN102209041B CN 102209041 B CN102209041 B CN 102209041B CN 201110196170 A CN201110196170 A CN 201110196170A CN 102209041 B CN102209041 B CN 102209041B
- Authority
- CN
- China
- Prior art keywords
- child node
- node
- state
- expectation
- resource
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Images
Landscapes
- Data Exchanges In Wide-Area Networks (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention discloses a scheduling method, device and system. The scheduling method comprises the followings steps of: receiving perceive information reported in real time by each sub node in a local area network by a main node; determining predetermined expectation states corresponding to each sub node; when the perceive information of a sub node does not meet the corresponding expectation state, regarding the sub nod as a to-be-processed sub node; generating a scheduling strategy; sending the scheduling strategy to the to-be-processed node. The main node determines state information of the nodes according to the perceive information reported by the sub nodes and judges whether a sub node is in an efficient and rational work state according to the expectation state predetermined for the sub node; if the current work state does not meet the predetermined expectation state, the scheduling strategy can be generated by using the state information according to the expectation state to schedule the task or resource of the sub node, so that the sub node can work in the expectation state, the resource can be more efficiently and reasonably used, and the resource waste in cloud computing is reduced.
Description
Technical field
The present invention relates to system for cloud computing field, relate in particular to a kind of dispatching method, Apparatus and system.
Background technology
The core concept of cloud computing, be by software by the computational resource unified management and the scheduling that connect with network in a large number, form a computational resource pond to user's on-demand service.The basic principle of cloud computing is by making Computation distribution in network on existing a large amount of distributed computer, but not in local computer or remote server, this makes the enterprise can be by resource switch to the application of needs, access computer and storage system according to demand.
In the concrete application of cloud computing, need enough resources, so that cloud computing can be tackled spike application and the multifarious application of network,, while producing the large-scale application of burst and the variation of the kind of application for specific resources, can meet user's application.
In order to meet user's application, tend to make resource distribution in cloud computing to apply required configuration than general routine and will exceed a lot, because a lot of resources are idle in routine is used, so the resource in cloud computing can cause serious waste.
Therefore, need the more rational dispatching method of one, business and resource are dispatched, to realize resource, more efficiently and reasonably utilize, reduce the wasting of resources in cloud computing.
Summary of the invention
In view of this, the invention provides a kind of dispatching method, Apparatus and system, its concrete scheme is as follows:
A kind of dispatching method, comprising:
The perception information of each child node real-time report in the network of host node receiving area, described perception information at least comprises: the business of child node or resource type, buty cycle and power consumption;
Definite corresponding with each child node, predefined expectation state, described expectation state at least comprises: the expectation power consumption of described each child node;
When the perception information of child node does not meet described expectation state, determine that described child node is pending child node;
With reference to the expectation state of described pending child node, and generate scheduling strategy according to the perception information of each child node real-time report;
Send described scheduling strategy to described pending child node, described scheduling strategy carries out the basis of task and scheduling of resource as described pending child node.
Preferably, also comprise:
The perception information of described each child node real-time report of storage.
Preferably, the described perception information of described foundation determines that the process of the state information of child node comprises:
According to the mode of protocal analysis, flow identification or payload detection identification, determine the type of service of each child node.
Preferably, described type of service is: voice, video, browse, search engine, multimedia, network game, animation, map generate fast, a kind of or several combination arbitrarily in science calculating, instant messaging, microblogging or ecommerce.
Preferably, the state information of described child node does not meet described expectation state and at least comprises: the power consumption in described state information is greater than the expectation power consumption of child node.
Preferably, the expectation state of the described pending child node of described reference, and comprise according to the process that the perception information of each child node real-time report generates scheduling strategy:
According to the buty cycle in described pending child node state information, determine that in described pending child node, resource status is that idle resource is target resource;
Scheduling strategy is turn-offed in the operation that generation comprises described target resource sign.
Preferably, described expectation state also comprises: the expectation traffic carrying capacity of child node.
Preferably, described state information does not meet described expectation state and at least comprises: the traffic carrying capacity of child node is greater than the expectation traffic carrying capacity of child node.
Preferably, the expectation state of the described pending child node of described reference, and comprise according to the process that the perception information of each child node real-time report generates scheduling strategy:
Determine the scheduling sublayer node of described pending child node;
Obtain the state information of scheduling sublayer node described in the state information of each child node of described storage;
Calculate the difference traffic carrying capacity of the traffic carrying capacity of described pending child node and the expectation traffic carrying capacity of described pending child node;
According to the buty cycle in the state information of described scheduling sublayer node, determine the idling-resource amount of described scheduling sublayer node;
When described idling-resource amount is not less than the required stock number of described difference traffic carrying capacity, generates the business that comprises described scheduling sublayer node identification and described difference traffic carrying capacity and shift scheduling strategy.
A kind of dispatching device, comprising:
Receiver module, for the perception information of each child node real-time report in host node reception network, described perception information at least comprises: the business of child node or resource type, buty cycle and power consumption;
Expectation state determination module, corresponding with each child node for determining, predefined expectation state, described expectation state at least comprises: the expectation power consumption of described each child node;
Pending node determination module, for when the state information of child node does not meet described expectation state, determines that described child node is pending child node;
Scheduling strategy generation module, for the expectation state with reference to described pending child node, and generates scheduling strategy according to the perception information of each child node real-time report;
Sending module, for sending described scheduling strategy to described pending child node, described scheduling strategy carries out the basis of task and scheduling of resource as described pending child node.
Preferably, also comprise: memory module, for storing the perception information of described each child node real-time report.
A kind of dispatching patcher, comprising: multi tiered computing resource, and every one deck computational resource comprises multiple Local Area Networks, each Local Area Network comprises: host node and multiple child node, be provided with dispatching device as above on described host node.
Dispatching method disclosed by the invention, the perception information that host node reports according to child node determines that whether the state information of node and the expectation state of setting for child node in advance carry out predicate node in an efficient and rational operating state, if current operating state does not meet predefined expectation state, can reference expected state, utilize state information to generate scheduling strategy, task or resource to node are dispatched, thereby make node can be operated in expectation state, realize resource and more efficiently and reasonably utilize, reduced the wasting of resources in cloud computing.
Accompanying drawing explanation
In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, to the accompanying drawing of required use in embodiment or description of the Prior Art be briefly described below, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skills, do not paying under the prerequisite of creative work, can also obtain according to these accompanying drawings other accompanying drawing.
Fig. 1 is the flow chart of the disclosed dispatching method of the embodiment of the present invention;
Fig. 2 is the expectation state of the described pending child node of the disclosed reference of the embodiment of the present invention, and according to the perception information of each child node real-time report, generates the flow chart of scheduling strategy;
Fig. 3 is the expectation state of the described pending child node of the disclosed another reference of the embodiment of the present invention, and according to the perception information of each child node real-time report, generates the flow chart of scheduling strategy;
Fig. 4 is level and the functional schematic of the disclosed DCSP agreement of the embodiment of the present invention;
Fig. 5 is the structural representation of the disclosed dispatching device of the embodiment of the present invention;
Fig. 6 is the structural representation of the disclosed dispatching patcher of the embodiment of the present invention.
Embodiment
Below in conjunction with the accompanying drawing in the embodiment of the present invention, the technical scheme in the embodiment of the present invention is clearly and completely described, obviously, described embodiment is only the present invention's part embodiment, rather than whole embodiment.Based on the embodiment in the present invention, those of ordinary skills, not making the every other embodiment obtaining under creative work prerequisite, belong to the scope of protection of the invention.
The embodiment of the invention discloses a kind of dispatching method, the method is based upon child node in network the perception information of himself is successfully reported on the basis of the host node that generates scheduling strategy.Because the core concept of cloud computing is by the computational resource unified management and the scheduling that connect with network in a large number by software, a large amount of computational resources can be divided into multilayer according to conditions such as handling property or regions, the computational resource of every one deck can be divided into multiple Local Area Networks equally, each Local Area Network comprises a host node and multiple child node, and the child node in high-rise computational resource in Local Area Network can be the host node of low layer Local Area Network.To be divided into example according to regional condition, suppose national resource to carry out unified management and scheduling, so can be according to province, city, counties etc. carry out layering, province is positioned at ground floor, city is positioned at the second layer, distribute downwards successively, every one deck all comprises multiple Local Area Networks, such as Local Area Network corresponding to Hebei province, the Local Area Network that Henan Province is corresponding, and in the Local Area Network of ground floor, can be divided into multiple second layer Local Area Networks according to prefecture-level city again, for example, for Hebei province's Local Area Network, can comprise: Shijiazhuang City region network, Baoding Local Area Network, region, Tangshan City network etc.A point involved host node in Shijiazhuang City region network, Baoding Local Area Network, region, Tangshan City network, whole resources in its region network are carried out to unified management and scheduling, and the host node of each second layer Local Area Network is as the child node of the host node of ground floor Local Area Network, finally, resource status in whole prefecture-level city's Local Area Networks is aggregated on the host node in corresponding Hebei province, by the host node in corresponding Hebei province, the resource in the whole province is carried out to unified management scheduling.
The embodiment of the invention discloses a kind of dispatching method, from the host node in each layer region network, the angle that manages scheduling from node in its region is described, its flow process as shown in Figure 1, comprising:
The perception information of each child node real-time report in host node receiving area network in step S11, each layer region network, described perception information at least comprises: the business of child node or resource type, buty cycle and power consumption;
Relevant perception information is collected and stored to child node, for example, the business of child node or resource type, buty cycle, power consumption etc., as perception information, report host node.Child node can be server, router, switch, bridge, intelligent HUB, PC, the network printer etc.Hardware resource type can be divided into server, router, switch, bridge, intelligent HUB, PC, the network printer, dish battle array, uninterrupted power supply UPS, rack (server cabinet, network cabinet, control desk rack), refrigeration plant, various kinds of sensors etc.
The type of application software resources can be classified according to type of service, and the type of systems soft ware resource can be according to operating system, various language processor and various data base management system 3 classes.
According to the mode of protocal analysis, flow identification or payload detection identification, determine the type of service of each child node.Described type of service is: voice, video, browse, search engine, multimedia, network game, animation, map generate fast, a kind of or several combination arbitrarily in science calculating, instant messaging, microblogging or ecommerce.In addition, can also carry out identification services type by modes such as agreement process analysis identification, connection mode identification, the identifications of stream statistic behavior.
Step S12, definite corresponding with each child node, predefined expectation state, described expectation state at least comprises: the expectation power consumption of described each child node;
System is set the state of each child node in advance, as expectation state, wherein at least comprises: the expectation power consumption of child node.Can also comprise the expectation traffic carrying capacity of child node etc.
Step S13, when the perception information of child node does not meet described expectation state, determine that described child node is pending child node;
The state information of concrete described child node does not meet described expectation state and at least comprises: the power consumption in described state information is greater than the expectation power consumption of child node.Show to carry out dispatch deal to this child node.
The expectation state of step S14, the described pending child node of reference, and generate scheduling strategy according to the perception information of each child node real-time report;
According in above-mentioned steps, when the power consumption in described state information is greater than the expectation power consumption of child node, the expectation state of the described pending child node of described reference, and the process that generates scheduling strategy according to the perception information of each child node real-time report is as shown in Figure 2, comprising:
Buty cycle in step S21, the described pending child node state information of foundation, determines that in described pending child node, resource status is that idle resource is target resource;
Power consumption is greater than expects that power consumption usual reason is, does not carry out the resource of work in opening, so determine after idle resource by buty cycle, idling-resource is processed.
Scheduling strategy is turn-offed in the operation that step S22, generation comprise described target resource sign.
It can be a special instruction generating that strategy is turn-offed in this operation, in instruction, comprise target resource sign, be used for determining which resource is dispatched, concrete form can be, directly sign is arranged in signaling for storing the region of target resource sign, hypothetical target resource identification is 5, at the region division of storage target resource sign, be directly 5, or, storage target resource identifies corresponding binary code, as 101, the present embodiment does not limit concrete form and only has above-mentioned two kinds, every can playing identifies to target resource the form representing, it is all the scope of the present embodiment protection.The present embodiment also can utilize existing SNMP (Simple Network Management Protocol, Simple Network Management Protocol), or XMPP (Extensible Messaging and Presence Protocol, scalable message Presence Protoco) thus signaling improve and obtain.
Step S15, send described scheduling strategy to described pending child node, described scheduling strategy carries out the basis of task and scheduling of resource as described pending child node.
Scheduling strategy is sent to pending child node, make its scheduling of carrying out task and resource according to this scheduling strategy, make its state information meet described expectation state.
Further, the present embodiment also comprises: the perception information of described each child node real-time report of storage.This process can occur in above-mentioned processing procedure when host node receives perception information, or afterwards.
The disclosed dispatching method of the present embodiment, perception according to node to information, obtain the perception information of node, and when perception information does not meet the expectation state of node, generate resource and the task of scheduling strategy to node and adjust, thereby node can be operated under predefined expectation state, improved the resource utilization of node, make node with maximum efficiency, to carry out work in real time, thereby reduced the wasting of resources in cloud computing.
Further, in above-described embodiment, expectation state also comprises: the expectation traffic carrying capacity of child node.Now, the situation that described state information does not meet described expectation state also comprises: the traffic carrying capacity of child node is greater than the expectation traffic carrying capacity of child node.It is the expectation traffic carrying capacity that task amount that child node is born is greater than predefined this node.Predefined expectation traffic carrying capacity is the traffic carrying capacity that this node can efficiently be processed, if be greater than this traffic carrying capacity, will inevitably cause the reduction of child node treatment effeciency.When the traffic carrying capacity of child node is greater than the expectation traffic carrying capacity of child node, with reference to the expectation state of described pending child node, and the process that generates scheduling strategy according to the perception information of each child node real-time report is as shown in Figure 3, comprising:
Step S31, determine the scheduling sublayer node of pending byte point;
In the present embodiment, determine that be that scheduling sublayer node is processed with pending child node apart from minimum child node.
Step S32, obtain the state information of scheduling sublayer node described in the state information of each child node of described storage;
According to the state information of each child node of storage, obtain the state information of scheduling sublayer node, the traffic carrying capacity of pending child node can be transferred on this scheduling sublayer node judging whether.
The difference traffic carrying capacity of the expectation traffic carrying capacity of step S33, the traffic carrying capacity of calculating described pending child node and described pending child node;
The traffic carrying capacity that obtains pending child node exceeds the part of expecting traffic carrying capacity.
Step S34, according to the buty cycle in the state information of described scheduling sublayer node, determine the idling-resource amount of described scheduling sublayer node;
Obtain the idling-resource amount of scheduling sublayer node.
Step S35, judge whether the required stock number of described difference traffic carrying capacity is less than the idling-resource amount of described scheduling sublayer node, if so, performs step S36, if not, finishes;
The business that step S36, generation comprise described scheduling sublayer node identification and described difference traffic carrying capacity shifts scheduling strategy;
Then business is shifted to scheduling strategy and send to pending child node, pending child node sends to difference traffic carrying capacity and identifies corresponding scheduling sublayer node, is exceeded the part of default traffic carrying capacity by scheduling sublayer node processing.
The present embodiment does not limit and will be defined as scheduling sublayer node with the nearest child node of pending child node, it can set other principle equally, or, by other child nodes in this layer region network according to sorting with the ascending of distance of pending child node, in the time cannot meeting the needs of difference traffic carrying capacity with pending child node apart from the idling-resource amount of minimum child node, according to ascending order, choose next son node, as scheduling sublayer node, then repeat said process.
The disclosed method of the present embodiment, by expecting that by exceeding child node the traffic carrying capacity of traffic carrying capacity is dispatched on scheduling node, makes child node its traffic carrying capacity can be maintained to desired value, realizes the efficient processing to business.This agreement is based upon on the basis of transmitting of lower-layer protocols.
Further, disclosed above-described embodiment dispatching method can be realized by agreement.Technical staff on this basis, according to above-mentioned thought, designed a kind of dynamic cooperation scheduling protocol (Dynamic Cooperative Scheduling Protocol, be called for short: DCSP), as business and resource coordinating perception and scheduling controlling rule and agreement.The function of this agreement, as shown in above-mentioned embodiment of the method, can realize the collaborative perception function of resource and utilance thereof, thereby can make the processing of business more professional and efficient, and the utilization of resource is more efficient with rationally.The level of the disclosed DCSP agreement of the embodiment of the present invention and function as shown in Figure 4, comprising: type of service layer, DCSP layer, TCP/UDP layer, IP layer, link layer and hardware layer.Its function comprises: scheduling feature, control function, service-aware function, resource perceptional function, energy consumption perceptional function, synergistic function, function for monitoring etc.The network packet that application DCSP agreement is carried out collaborative perception and scheduling controlling management is containing four fundamentals: network element, agency, management system and DCSP agreement.A network element is a network node comprising in DCSP agency the network management region in being managed, for example child node in above-described embodiment.Relevant perception information is collected and stored to network element, and use DCSP to make these information useful to management system.Network element can be server, router, switch, bridge, intelligent HUB, PC, the network printer etc.Agency is a network management software module in network element.The pre-concern information of agency in can sense correlation region, and can be converted into the consistent form of DCSP.Management system executive utility monitor network element.Management system for management and scheduling corresponding processing and storage resources are provided.In any one network management region, at least there are two management systems, so that redundancy to be provided, prevent fault.
If realize above-mentioned functions, need to realize by corresponding order, for example, the perception information of perception self, can, by reading (get), write (set), notice (inf) and traversal (Traversal) operational order realize.Management system is carried out sensing network element with read command, checks the different variable of network elements maintain.Management system is used write order control network element, changes the variate-value being stored in network element.Network element is used inf order to management system reporting event asynchronously.When the event of certain type occurs, network element sends a notice to management system.Management system is used traversal operation to determine which variable network element supports, thereby collects relevant information.
With existing protocol massages structural similarity, DCSP message comprises multiple territories equally, type of message indicates the carrying IP address of DCSP, business and resource type code, business, the perception propelling movement address of business and the details of business and resource etc., and by obtaining operation get-request, extract one or more parameter values; Setting operation set-request, arranges one or more parameter values; Operation response get-response, the one or more parameter values that return, it is the operation response of two kinds of operations above.And operate inf by notice, notify some event or behavior to occur.Traversing operation traversal, the types of variables of supporting according to network element, collection information is in corresponding argument table.
In embodiment disclosed by the invention, the realization of DCSP agreement can be used for reference the implementation of SNMP, by configuration file, provides configuration information, and sets up business and resource information bank, thereby realizes this agreement.
A DCSP message is comprised of three parts, i.e. public DCSP stem, get/set stem inf/traversal stem, variable binding.This sentences public DCSP stem and get/set stem is that example is introduced.
(1) public DCSP stem
Totally three fields: version, cloud cluster body (community) and type.
What write version field is that version number subtracts 1, for DCSP (being DCSPV1), should write 0.
Cloud cluster body is exactly a character string, and as the plaintext password between managing process and agent process, conventional is 6 characters " public ".
According to type, insert a numeral in 0~4, its corresponding relation intention as shown in table 1.
Table 1: type
Type | Title |
0 | get-request |
1 | get-response |
2 | set-request |
3 | inf |
4 | traversal |
(2) get/set stem
Request identifier (request ID)
This is the integer value being arranged by managing process.Agent process also will return to this request identifier when sending get-response message.Managing process can send get message to many agencies simultaneously, and these messages are all used UDP to transmit, and what first send likely arrives afterwards.Be provided with response message that request identifier can make managing process identify to return for which request message.
Error state (error status)
While being answered by agent process, insert a numeral in 0~5, in Table 2 description.
Table 2: error state is described
Error index (error index)
When there is the mistake of noSuchName, badValue or readOnly, the integer being arranged when answering by agent process, the skew of the variable that it indicates mistake in variable list.
Or the DCSP agreement in this enforcement can be expanded, realize child node and report in real time perception information with reference to XMPP (Extensible Messaging and Presence Protocol, scalable message Presence Protoco) protocol family.
XMPP protocol family adopts order and the data in XML language definition agreement, it has utilized XML highly versatile, easily expansion and machine-readable advantage, has facilitated the expansion of agreement in function and the intercommunication with other application thereof.In addition, with the order of the XML language definition of textual form, also avoided to a certain extent the introducing of binary system viral code.
Further, the present invention discloses the dispatching device of the above-mentioned dispatching method of application, its structure as shown in Figure 5, comprising: receiver module 51, expectation state determination module 52, pending node determination module 53, scheduling strategy generation module 54 and sending module 55.Wherein:
Further, also comprise: memory module 56, for storing the perception information of described each child node real-time report.
In the disclosed dispatching device of the present embodiment, the course of work of modules, as described in above-mentioned embodiment of the method, does not repeat them here.
Further, the present invention discloses a kind of dispatching patcher, its structure as shown in Figure 6, comprise multi tiered computing resource, the computational resource of every one deck is divided into multiple Local Area Networks 60, each Local Area Network comprises a host node and multiple child node, and in high-rise computational resource, the child node in Local Area Network can be the host node of low layer Local Area Network.On host node 61, be provided with dispatching device 62, described host node utilizes described dispatching device, the perception information of each child node 63 real-time report in the network of receiving area, described perception information at least comprises: the business of child node or resource type, buty cycle and power consumption, determine corresponding with each child node, predefined expectation state, described expectation state at least comprises: the expectation power consumption of described each child node, when the perception information of child node does not meet described expectation state, determine that described child node is pending child node, with reference to the expectation state of described pending child node, and generate scheduling strategy according to the perception information of each child node real-time report, send described scheduling strategy to described pending child node, described scheduling strategy carries out the basis of task and scheduling of resource as described pending child node.
The disclosed dispatching patcher of the present embodiment, the perception information that host node reports according to child node determines that whether the state information of node and the expectation state of setting for child node in advance carry out predicate node in an efficient and rational operating state, if current operating state does not meet predefined expectation state, can reference expected state, utilize state information to generate scheduling strategy, task or resource to node are dispatched, thereby make node can be operated in expectation state, realize resource and more efficiently and reasonably utilize, reduced the wasting of resources in cloud computing.
In this specification, each embodiment adopts the mode of going forward one by one to describe, and each embodiment stresses is and the difference of other embodiment, between each embodiment identical similar part mutually referring to.For the disclosed device of embodiment, because it corresponds to the method disclosed in Example, so description is fairly simple, relevant part illustrates referring to method part.
To the above-mentioned explanation of the disclosed embodiments, make professional and technical personnel in the field can realize or use the present invention.To the multiple modification of these embodiment, will be apparent for those skilled in the art, General Principle as defined herein can, in the situation that not departing from the spirit or scope of the present invention, realize in other embodiments.Therefore, the present invention will can not be restricted to these embodiment shown in this article, but will meet the widest scope consistent with principle disclosed herein and features of novelty.
Claims (12)
1. a dispatching method, is characterized in that, comprising:
The perception information of each child node real-time report in the network of host node receiving area, described perception information at least comprises: the business of child node or resource type, buty cycle and power consumption;
Definite corresponding with each child node, predefined expectation state, described expectation state at least comprises: the expectation power consumption of described each child node;
The perception information reporting according to child node determines that whether the state information of node and the expectation state of setting for child node in advance carry out predicate node in an efficient and rational operating state;
When the perception information of child node does not meet described expectation state, when the current operating state of child node does not meet predefined expectation state, determine that described child node is pending child node;
With reference to the expectation state of described pending child node, and generate scheduling strategy according to the perception information of each child node real-time report;
Send described scheduling strategy to described pending child node, task or resource to node are dispatched, and described scheduling strategy carries out the basis of task and scheduling of resource as described pending child node.
2. method according to claim 1, is characterized in that, also comprises:
The perception information of described each child node real-time report of storage.
3. method according to claim 2, is characterized in that, the described perception information of described foundation determines that the process of the state information of child node comprises:
According to the mode of protocal analysis, flow identification or payload detection identification, determine the type of service of each child node.
4. method according to claim 2, it is characterized in that, described type of service is: voice, video, browse, search engine, multimedia, network game, animation, map generate fast, a kind of or several combination arbitrarily in science calculating, instant messaging, microblogging or ecommerce.
5. method according to claim 1, is characterized in that, the state information of described child node does not meet described expectation state and at least comprises: the power consumption in described state information is greater than the expectation power consumption of child node.
6. method according to claim 5, is characterized in that, the expectation state of the described pending child node of described reference, and comprise according to the process that the perception information of each child node real-time report generates scheduling strategy:
According to the buty cycle in described pending child node state information, determine that in described pending child node, resource status is that idle resource is target resource;
Scheduling strategy is turn-offed in the operation that generation comprises described target resource sign.
7. method according to claim 2, is characterized in that, described expectation state also comprises: the expectation traffic carrying capacity of child node.
8. method according to claim 7, is characterized in that, described state information does not meet described expectation state and at least comprises: the traffic carrying capacity of child node is greater than the expectation traffic carrying capacity of child node.
9. method according to claim 8, is characterized in that, the expectation state of the described pending child node of described reference, and comprise according to the process that the perception information of each child node real-time report generates scheduling strategy:
Determine the scheduling sublayer node of described pending child node;
Obtain the state information of scheduling sublayer node described in the state information of each child node of described storage;
Calculate the difference traffic carrying capacity of the traffic carrying capacity of described pending child node and the expectation traffic carrying capacity of described pending child node;
According to the buty cycle in the state information of described scheduling sublayer node, determine the idling-resource amount of described scheduling sublayer node;
When described idling-resource amount is not less than the required stock number of described difference traffic carrying capacity, generates the business that comprises described scheduling sublayer node identification and described difference traffic carrying capacity and shift scheduling strategy.
10. a dispatching device, is characterized in that, comprising:
Receiver module, for the perception information of each child node real-time report in host node reception network, described perception information at least comprises: the business of child node or resource type, buty cycle and power consumption;
Expectation state determination module, corresponding with each child node for determining, predefined expectation state, described expectation state at least comprises: the expectation power consumption of described each child node;
Pending node determination module, for the perception information reporting according to child node, determine that whether the state information of node and the expectation state of setting for child node in advance carry out predicate node in an efficient and rational operating state, when the state information of child node does not meet described expectation state, be the current operating state of child node while not meeting predefined expectation state, determine that described child node is pending child node;
Scheduling strategy generation module, for the expectation state with reference to described pending child node, and generates scheduling strategy according to the perception information of each child node real-time report;
Sending module, for sending described scheduling strategy to described pending child node, task or resource to node are dispatched, and described scheduling strategy carries out the basis of task and scheduling of resource as described pending child node.
11. devices according to claim 10, is characterized in that, also comprise: memory module, and for storing the perception information of described each child node real-time report.
12. 1 kinds of dispatching patchers, it is characterized in that, comprising: multi tiered computing resource, every one deck computational resource comprises multiple Local Area Networks, each Local Area Network comprises: host node and multiple child node, be provided with the dispatching device as described in claim 10 or 11 on described host node.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201110196170.8A CN102209041B (en) | 2011-07-13 | 2011-07-13 | Scheduling method, device and system |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201110196170.8A CN102209041B (en) | 2011-07-13 | 2011-07-13 | Scheduling method, device and system |
Publications (2)
Publication Number | Publication Date |
---|---|
CN102209041A CN102209041A (en) | 2011-10-05 |
CN102209041B true CN102209041B (en) | 2014-05-07 |
Family
ID=44697710
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201110196170.8A Active CN102209041B (en) | 2011-07-13 | 2011-07-13 | Scheduling method, device and system |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN102209041B (en) |
Families Citing this family (15)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102520785B (en) * | 2011-12-27 | 2015-04-15 | 东软集团股份有限公司 | Energy consumption management method and system for cloud data center |
EP3089455B1 (en) * | 2013-04-06 | 2021-03-10 | Grass Valley Canada | Systems and methods for media distribution and management |
CN104216344B (en) * | 2013-06-03 | 2018-05-11 | 富泰华工业(深圳)有限公司 | Produce arranging system and method |
CN104281627B (en) * | 2013-07-12 | 2018-12-21 | 阿里巴巴集团控股有限公司 | A kind of method of data reporting processing result, system and track of issues device |
CN103336756B (en) * | 2013-07-19 | 2016-01-27 | 中国人民解放军信息工程大学 | A kind of generating apparatus of data computational node |
CN103458052B (en) * | 2013-09-16 | 2017-05-24 | 北京搜狐新媒体信息技术有限公司 | Resource scheduling method and device based on IaaS cloud platform |
CN104717247A (en) * | 2013-12-13 | 2015-06-17 | 中国电信股份有限公司 | Method and system for dynamically scheduling storage resources in cloud storage system |
CN103731495B (en) * | 2013-12-31 | 2017-06-23 | 曙光云计算技术有限公司 | The extension of PAAS platforms and the method and system of the resource quantity of contraction resource pool |
CN103685569A (en) * | 2013-12-31 | 2014-03-26 | 曙光云计算技术有限公司 | Method and system for automatically adjusting resource pool |
CN104504137A (en) * | 2014-12-31 | 2015-04-08 | 深圳市科漫达智能管理科技有限公司 | Data storage method and system |
CN104581192B (en) * | 2014-12-31 | 2018-01-30 | 江苏中科梦兰电子科技有限公司 | A kind of low delay based on sandwich construction, high concurrent media flow distribution method |
CN106209944A (en) * | 2015-05-07 | 2016-12-07 | 中兴通讯股份有限公司 | The resource apparatus management method of a kind of cloud service system and device |
CN106156631B (en) * | 2015-06-01 | 2019-03-12 | 上海红神信息技术有限公司 | A kind of service function and the uncertain software and hardware device of structural characterization corresponding relationship |
CN111338580B (en) * | 2020-02-29 | 2021-12-21 | 苏州浪潮智能科技有限公司 | Method and equipment for optimizing disk performance |
CN113992758B (en) * | 2021-12-27 | 2022-04-19 | 杭州金线连科技有限公司 | Dynamic scheduling method and device for system data resources and electronic equipment |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101123534B (en) * | 2007-09-29 | 2010-09-01 | 华中科技大学 | Network policy architecture for legal monitoring system and its policy processing method |
CN101925111A (en) * | 2009-06-16 | 2010-12-22 | 中兴通讯股份有限公司 | Method and device for determining scheduling strategy |
CN101997746A (en) * | 2009-12-18 | 2011-03-30 | 中国科学院声学研究所 | Dynamic service quality strategy method and system of home network |
CN102073546A (en) * | 2010-12-13 | 2011-05-25 | 北京航空航天大学 | Task-dynamic dispatching method under distributed computation mode in cloud computing environment |
Family Cites Families (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
KR101613229B1 (en) * | 2008-08-11 | 2016-04-19 | 코닌클리케 필립스 엔.브이. | A method for scheduling transmissions of global beacons in body area networks |
-
2011
- 2011-07-13 CN CN201110196170.8A patent/CN102209041B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101123534B (en) * | 2007-09-29 | 2010-09-01 | 华中科技大学 | Network policy architecture for legal monitoring system and its policy processing method |
CN101925111A (en) * | 2009-06-16 | 2010-12-22 | 中兴通讯股份有限公司 | Method and device for determining scheduling strategy |
CN101997746A (en) * | 2009-12-18 | 2011-03-30 | 中国科学院声学研究所 | Dynamic service quality strategy method and system of home network |
CN102073546A (en) * | 2010-12-13 | 2011-05-25 | 北京航空航天大学 | Task-dynamic dispatching method under distributed computation mode in cloud computing environment |
Also Published As
Publication number | Publication date |
---|---|
CN102209041A (en) | 2011-10-05 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN102209041B (en) | Scheduling method, device and system | |
Qiu et al. | A novel energy-aware fault tolerance mechanism for wireless sensor networks | |
CN103036736B (en) | A kind of configuration equipment monitoring system based on data source and method | |
CN101924650B (en) | Method for implementing services and intelligent server autonomy of failure information system | |
CN109788315A (en) | Video transcoding method, apparatus and system | |
CN109213792A (en) | Method, server-side, client, device and the readable storage medium storing program for executing of data processing | |
CN107317707B (en) | SDN network topology management method based on point coverage set | |
CN102014403A (en) | Method and system for transmitting network topology information | |
CN111885439B (en) | Optical network integrated management and duty management system | |
CN103581307A (en) | Publishing/subscribing system based on clusters and method for guaranteeing reliability of publishing/subscribing system based on clusters | |
CN107210607A (en) | Method and system for providing energy service | |
CN107104961A (en) | Distributed real-time video monitoring processing system based on ZooKeeper | |
CN103259701A (en) | Message bus implementation method for complex production process management system | |
CN102984012A (en) | Management method and system for service resources | |
CN102088475B (en) | System and method for executing combined service with centralized control flow and distributed data flow | |
Zhao et al. | PLOFR: An online flow route framework for power saving and load balance in SDN | |
CN103763353A (en) | Water conservation data exchange model and method | |
CN103179009B (en) | A kind of dynamic self-adapting call method of distributed management system | |
CN101296232B (en) | Adapting device and method with multi-network management and multi-north interface | |
GuiLing et al. | Service-oriented grid architecture and middleware technologies for collaborative e-learning | |
WO2021148015A1 (en) | Computer system | |
JP2010527475A (en) | Information appliance data storage system and data processing method | |
CN111294553B (en) | Method, device, equipment and storage medium for processing video monitoring service signaling | |
CN112765294A (en) | Meteorological big data processing and scheduling system | |
CN209313868U (en) | A kind of distribution cloud stocking system |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant |